Abstract Cells are highly complex living nanomachines with beautiful structures of great precision. This is true not only for free living organisms like ciliates or radiolarians, but also for cells inside the human body. These complicated structures are directly linked to the physiological functions of cells, and alterations in cell geometry are a hallmark of many disease states. Yet in most cases we have almost no information about how cells determine their geometry at the level of organelle size and shape. Thus, understanding the origins of cell geometry remains a fundamental unsolved problem in cell biology. Part of the challenge is that cell geometry involves multiple spatial scales ranging from molecules up to the whole cell. Spanning this gap between scales requires us to go beyond traditional molecular biology approaches and bring in methods from physics and engineering. For this reason my proposal is based on an integrated combination of approaches, using several different model organisms and cell types to address the origins of cell geometry at several different size scales. At the level of single organelles, I will continue to probe the mechanism of flagellar length control as a paradigm for organelle size regulation, with a focus on using quantitative methods to test a series of mechanistic models for how a cell might be able to sense the length of its flagellum. At the same time, we will apply the lessons and approaches that we have developed for thinking about flagella to examine size control and geometry of other cellular organelles, singly and in combination. By considering multiple organelles at the same time, we can learn how to view cell geometry at a more integrative level. At a larger scale, we will continue our development of the classic model organism, Stentor coeruleus, as a genomic model system for analyzing global cell morphogenesis and regeneration. Using Stentor, we intend to pursue the two linked questions of how a cell knows that is geometry has been perturbed, and how it directs the re-assembly of a correct cell geometry, both questions that have general significance to all cell types but which are particularly easy to study in Stentor. Our proposed work is unified by the focus on a single question ? where does geometry come from inside a cell. We will use different model systems to address different aspects of this question, but in all cases we will take an interdisciplinary approach that combines tools of genetics, genomics, microscopy, image analysis, and mathematical modeling.